Thread by @jeremyfaust

📢2020 will be the deadliest year in modern American history.

We will see >3 million deaths for the first time ever.

In fact, well over that number.

Join me for a thread on basic epidemiology and how we count deaths and where things stack up, historically.

🧵 ...

We'll probably see 3.15 - 3.25 million deaths this year (from all causes).

That is around 300,000 or more will be more than was "expected"

But is that normal variance?

Is that just a "bad year" combined with a bigger population?

No and no.

How do we keep track of deaths in this country?

It's easy.

The CDC counts all the deaths, and divides by the population.

(We're not talking about deciding what caused the deaths. This is ALL CAUSES).

We express deaths as "deaths per 100,000 people."

So how are we looking?

In 2018, there were 867 deaths per 100,000 people, or 8.67 for every 1,000.

The number can vary (see below).

As you can see, we were on track to have around 2.9 million deaths in 2020, and somewhere around 870 and closer to 890 deaths per 100,000, if we had a really bad year.

With 250,000+ excess deaths already recorded by CDC and others, it is reasonable to guess that we'll be closer to 300,000 by Dec 31.

With over 330 million Americans, that's 3.2 million deaths, and a crude rate of 970 deaths per 100,000.

970 deaths per 100,000 is *a lot*.

That's around a 10% increase in all cause deaths.

A 10% increase in all deaths in *one year*

Usually, as per above, you might see a 1-2% change in either direction over a couple of years.

To have a 10% increase in all-cause deaths is truly impressive (and not in a good way)

Please note that there was *not* an increase in all-cause deaths around 2009 H1N1.

In fact, we had decent years in the US.

Some may argue that our rates would've been EVEN better those years but for H1N1.

Perhaps.

But whatever the effect, it didn't "overwhelm" other trends.

I mention this because I am *not* an alarmist by habit or disposition. I do not "fail to notice" when there's actually not a huge problem.

I tend to be fairly nihilistic about many things.

But with COVID-19, I can't ignore the data.

This is literally killing at historic rates.

In fact, to get to 970 deaths per 100,000 Americans, you have to go back to the 1960s!

Since then, we've added seatbelts, better cancer treatments, life-saving cardiac procedures.

So for us to be recording the same crude death rates as there were during the 1960s because of COVID-19 is LITERALLY backtracking on generations of progress.

In fact, @CarlosdelRio7, Zhenqiu Lin and I found it was more dangerous to be alive in NYC in 2020 than it was there in the early 1910s (!)

And when we compare COVID in Spring in NYC to 1918 H1N1 in Fall (the peak)? Covid=70% as bad, despite modern ICUs.
jamanetwork.com/journals/jaman…

All of this would be sad but useless information, if we were powerless to do anything.

But we are not.

If a vaccine rolls out in 2021, lives will be saved.

Wearing masks and limiting gatherings will keep us all alive long enough to get that vaccine.

In sum, the death rates we are seeing are truly historic.

I say this as someone who usually isn't alarmed by smaller (though epidemiologically important) phenomena.

We *can* influence how many live and die as a result of this.

Saving lives is *patriotic*

Let's do this.